Label Switched Path (LSP) Data Path Delay Metrics in Generalized MPLS and MPLS Traffic Engineering (MPLS-TE) Networks
RFC 6777

Note: This ballot was opened for revision 06 and is now closed.

(Stewart Bryant) Yes

(Adrian Farrel) Yes

(Ron Bonica) No Objection

(Wesley Eddy) (was Discuss) No Objection

Comment (2012-08-30 for -07)
No email
send info
I agree with the many others that have commented on confusing usage of "applications".

(Stephen Farrell) No Objection

Comment (2012-08-28 for -07)
No email
send info
- Is this really for general application developers as the abstract
implies? Seems odd. How'd they know that MPLS was even in use?  I'm
sure you know who this is intended for, but it wasn't clear to this
reader.

- p4: Could you add a reference for UNI? Maybe rfc 4208?

- p4, typo: s/complements/complement/

- 5.4 and elsewhere: I don't get how the metric can be ms or
undefined. If you tell me RRFD=20 how'd I know if that's ms or
bananas? That's probably just my ignorance of metrics, but if this
were intended for general application developers, I suspect they'd
have the same problem.

- p23, very nice typo: s/potion/portion/ but maybe leave it in:-)

- section 12: perhaps you should say that applications ought not trust
in these statistics too much as they could be fooled.

(Brian Haberman) No Objection

Comment (2012-08-29 for -07)
No email
send info
I agree with Stephen's comment about the oddness of having "application developers" mentioned in the abstract.  I can't see them really caring if MPLS, Ethernet, or carrier pigeons are used to move their data as long as the data is moving.

(Russ Housley) No Objection

Barry Leiba No Objection

Comment (2012-08-29 for -07)
No email
send info
About the "application" discussion, this is copied from my email on the topic:
----------------------
Brian says:
>> I agree with Stephen's comment about the oddness of having
>> "application developers" mentioned in the abstract.  I can't
>> see them really caring if MPLS, Ethernet, or carrier pigeons
>> are used to move their data as long as the data is moving.

Adrian says:
> What an application is depends where you are in the stack,
> doesn't it?
>
> An application in this context is the thing what uses an LSP.
> Can you think of a better term?  "User" has sometimes been used,
> but I hate to think of you expecting to find a human using an LSP.

Barry says:
Why not "LSP application", throughout?  And a definition could be put
into Section 2 if someone thinks it's needed:

'In this document, "LSP application" refers to the thing what uses LSP.'

...or something more explanatory and less flippant.
----------------------

(Pete Resnick) No Objection

Comment (2012-08-29 for -07)
No email
send info
I agree with Brian and Stephen regarding "applications": As we use that word in the IETF, I can't imagine applications using this technology (or anything having to do with MPLS). If you mean something specific by "applications", you should probably explain that.

(Robert Sparks) No Objection

Comment (2012-08-28 for -07)
No email
send info
Why does the document describe units as a _real_ number of milliseconds? (π milliseconds doesn't seem useful).
Are you trying to discourage implementations from quantizing on whole milliseconds?
If so, how much sub-millisecond precision do you want to maintain?

(Martin Stiemerling) No Objection

(Sean Turner) No Objection